First-time adverse drug reactions, survival analysis, and the share of adverse drug reactions in treatment discontinuation in real-world rheumatoid arthritis patients: a comparison of first-time treatment with adalimumab and etanercept.

BACKGROUND: This study aims to compare nature and frequency of adverse drug reactions (ADRs), time to first ADR, drug survival, and the share of ADRs in treatment discontinuation of first-time treatment with adalimumab (ADA) and etanercept (ETN) in real-world RA patients. RESEARCH DESIGN AND METHODS: Retrospective, single-center cohort study including naïve patients treated between January 2003-April 2020. Time to first ADR and drug survival of first-time treatment were studied using Kaplan-Meier and Cox-regression models up to 10 years, with 2- and 5-year post-hoc sensitivity analysis. Nature and frequencies of first-time ADRs and causes of treatment discontinuation were assessed. RESULTS: In total, 416 patients (ADA: 255, ETN: 161, 4865 patient years) were included, of which 92 (22.1%) experienced ADR(s) (ADA: 59, 23.1%; ETN: 33, 20.4%). Adjusted for age, gender and concomitant conventional DMARD use, ADA was more likely to be discontinued than ETN up to 2-, 5- and 10-year follow-up (adjusted HRs 1.63; 1.62; 1.59 (all p<0.001)). ADRs were the second reason of treatment discontinuation (ADA 20.7%, ETN 21.4%). CONCLUSIONS: Despite seemingly different nature and frequencies, ADRs are the second reason of treatment discontinuation for both bDMARDs. Furthermore, 2-, 5-, and 10-year drug survival is longer for ETN compared to ADA.

Predicting protein condensate formation using machine learning.

Membraneless organelles are liquid condensates, which form through liquid-liquid phase separation. Recent advances show that phase separation is essential for cellular homeostasis by regulating basic cellular processes, including transcription and signal transduction. The reported number of proteins with the capacity to mediate protein phase separation (PPS) is continuously growing. While computational tools for predicting PPS have been developed, obtaining a proteome-wide overview of PPS probabilities has remained challenging. Here, we present a phase separation analysis and prediction (PSAP) machine-learning classifier that, based solely on the amino acid content of a training set of known PPS proteins, can determine the phase separation likelihood for each protein in a given proteome. Through comparison with PPS databases, existing predictors, and experimental evidence, we demonstrate the validity and advantages of the PSAP classifier. We anticipate that the PSAP predictor provides a useful tool for future research aimed at identifying phase separating proteins in health and disease.